Image Optimization Queue with Generative AI: A 2027 Project Guide

A convincing Generative AI portfolio project explains the result as well as showing it. Here, a student web-performance learner asks: How much verified file-size reduction is recorded by image group? You will build an evidence-grounded assistant prototype and follow the relevant records from a small fixture to an independently checked answer.

This is a 2027 study-edition project brief. All example records and operational situations are fictional. Start with basic Generative AI knowledge, the ability to inspect structured data and a local practice environment. The complete first version should answer the stated question before adding accounts, real integrations or additional screens.

Define the project rule

Saved size counts only verified conversions after a visual check. A smaller file is not automatically acceptable if text, transparency or proportions are damaged. For this exercise, one row represents one image_job. The stable key is record_id; image_group is the grouping or filtering label. The numeric saved_kb field uses kilobytes. Accepted states are queued, verified, rejected.

The summary selects status verified and applies sum to the selected numeric values. Keep this definition visible in the interface, response or report so another learner can reproduce it. Use positive finite values in the starter fixture and define any future zero or missing-value policy explicitly.

Choose the Generative AI approach

Use the application data as evidence and keep calculations in deterministic code. Ask a language model to explain a supplied result rather than inventing the result from an instruction. Treat record text as untrusted content, make unsupported questions visible, and evaluate against a small human-written answer set before describing the assistant as useful.

The three design decisions to explain in your project notes are:

• Deterministic calculations outside the model. Apply this choice to image_job records and explain its effect on saved_kb.
• Evidence references in answers. Apply this choice to image_job records and explain its effect on saved_kb.
• Explicit abstention when data is absent. Apply this choice to image_job records and explain its effect on saved_kb.

Work through a small sample

The three records below are deliberately small enough to inspect by hand. They are a compact specification for the data shape and expected result, and can later be expanded into a larger synthetic fixture.

[
  {
    "record_id": "image-optimization-01",
    "image_group": "Homepage Banners",
    "saved_kb": 320,
    "status": "verified",
    "recorded_on": "2027-01-04"
  },
  {
    "record_id": "image-optimization-02",
    "image_group": "Course Cards",
    "saved_kb": 480,
    "status": "queued",
    "recorded_on": "2027-01-05"
  },
  {
    "record_id": "image-optimization-03",
    "image_group": "Blog Illustrations",
    "saved_kb": 160,
    "status": "verified",
    "recorded_on": "2027-01-06"
  }
]

Only the first and third records match verified. Their saved_kb values are 320 and 160; the full-fixture result is 480 kilobytes, with 2 matching records. Selecting image_group = Homepage Banners leaves the first record, so the result becomes 320 kilobytes, with 1 match. This provides two independent checks for the filter and aggregation.

The full-fixture output contract is:

{
  "matching_count": 2,
  "operation": "sum",
  "value": 480,
  "unit": "kilobytes"
}

Check the result by image_group

The overall figure answers the all-records version of the question. Grouping by image_group produces the following separate expectations, using the same verified rule. The middle group has a record in the source fixture but no qualifying observation.

Use this breakdown when answering the group-level question. Compare like-for-like units and keep the matching count visible, especially for averages. A different filter changes which records enter the calculation; it does not change the meaning of the metric.

The code blocks describe sample data and an expected result. Implement the surrounding feature using the installed version of your chosen tools, then save the actual run evidence separately from these expectations.

Build the feature in three passes

1. Calculate before asking for prose

Use ordinary code to select verified image_job records and calculate the sum of saved_kb. Pass the verified result, units and matching record identifiers to the assistant. The model should explain supplied evidence rather than invent a numerical answer.

2. Define the answer and abstention format

Ask for a brief answer to the image_group question with references to the supplied records. Include an explicit outcome for absent or inconsistent data. Treat labels and other record text as data, including any instruction-like text inside a fictional record.

3. Evaluate a small answer set

Write expected answers for the full fixture, one image_group filter, no matching records and an unsupported question. Compare numbers deterministically and review evidence references separately. Record observed failures before adding more prompts or calling the prototype reliable.

Find the mistake before polishing the result

A useful defect to introduce deliberately is: Rejected conversions are included in claimed size savings. First reproduce it against the fixture, then identify the smallest rule or state transition that explains the mismatch. Fix that boundary and rerun the full-fixture and selected-label examples. This creates a stronger explanation than changing several unrelated parts until the screen looks right.

The scenario-specific acceptance check is: Reject a damaged conversion and remove its saving from the verified total. Preserve before-and-after evidence. Add a missing image_group, a negative saved_kb, an unrecognised status and a duplicate record_id as separate invalid cases. Define the rejection or duplicate-handling policy before testing, and verify that invalid input does not silently change valid existing records.

Check the Generative AI concerns as well:

• The answer must not invent a missing record.
• Text inside a record must not override the task.
• A numerical answer must match the verified calculation.

For an empty filter match, the expected sum value is 0. Keep matching_count at zero. This distinction prevents an empty dataset from being interpreted as a measured observation.

Explain the finished work in a portfolio

Show the student web-performance learner question, the record definition and the expected calculation first. Then demonstrate a image_group filter and the acceptance check. Include the fixture, setup instructions, chosen tool versions, actual test output and one remaining limitation in the repository. If you show a screenshot, pair it with the input that produced it. Another learner should be able to repeat the same result from a clean checkout.

For a focused extension, add a second image_group selection and compare the two groups using the same rule. Specify how missing groups appear and keep the original fixture as a regression case. Expand scope only after the current calculation, state handling and recovery path remain understandable.

Questions learners often ask

Which image_job records belong in the result?

Select records whose status is exactly verified and that satisfy the chosen image_group filter. Other states remain in the source data and can have separate summaries. Follow the domain rule above when a record changes state.

What should happen when no records match?

Return a zero total with a matching count of zero and a clear empty-result message. Keep the applied filter visible so the user can change it. Avoid using a loading indicator for a completed request that simply has no matching data.

How can I prove the saved_kb result is correct?

Reconcile the first and third sample records, then repeat the calculation for Homepage Banners alone. Save both expected and actual outputs, and rerun the scenario-specific acceptance check after making a change.

Continue learning with Groot Academy

Use this exercise as a concrete discussion point when choosing a learning path. Browse the current Groot Academy course directory and ask about the current syllabus, learning mode and project practice. Course availability and batch details should be confirmed directly with the academy.

For implementation details, consult the official Generative AI documentation matching the tools you install. This project defines an original practice brief; it does not depend on a forecast about 2027 software releases.

Continue with Educational Content Review using Generative AI, or compare Image Optimization Queue using Python. These companion practice guides are part of the 2027 learning collection.